API test template
My goal is to create a large number of API tests. I'm using UTF to design/create and ALM to manage them. The workflow is the same for all of them, but parameters are different. I'm looking for a easiest way to do that.
Most obvious solutuion is to have one test, that is driven by data table. But this option doesn't work for me as different iterations should be mapped to different requirements. So it would be still different tests.
I created a test in UTF that can be used as a template with set of input parameters. I was able to copy/paste this test in ALM to create more such tests. Input parameters are displayed fine in Parameters tab. But when I execute these tests, input parameters are blank, not even default value is used.
Is there a way to set input parameters for UFT created API test within ALM and hopefully with minimal actions?
Thanks in advance
Re: API test template
Crap I had written a long answer but it got lost since I needed to register. Anyway, I have written a framework around API test in UFT where my thought behind it was
a) get around all the limitations I found using API test
b) be able to create a clean report file that anyone can understand
c) make it possible for anyone to create tests without having to know UFT
So to be able to do this I used a GUI test as template test that will be copied to create new API test. There are several reasons for using a GUI test the main ones are
1) Its possible to customize the report
2) Its possible to integrate with ALM in a much much easier way
Since I use one template test that is copied I dont want to end up in maintenance hell so I use function libraries (basicly the template only calls a function).
As for test data I have there sources:
1) Global where stuff like environment url, username,password etc are stored
2) Attached to each test set/test I have test specific data which tells my template script what I actually want to test
3) And lastly my template looks at a custom field on the test set to determine which environment to run. This is vital to make sure that you can have the same test in several testsets scheduled to run without having to modify your excel each time
All in all this gives me this nice report that I think anyone can understand. You can expand/collapse all environments that are running, or iterations. And when you have a verification you get a nice expected/actual printout. Much nicer than getting Action Iteration 1, 2,etc I would say
All this being said, its not something you setup during coffee break but a good framework is a good investment.